METTL3 promotes colorectal carcinoma progression by regulating the m6A–CRB3–Hippo axis

Thirty CRC, 30 adenoma, and 30 adjacent normal (normal) tissues were obtained during surgery from Longhua Hospital, which is affiliated with the Shanghai University of Traditional Chinese Medicine. The diagnosis of CRC and adenoma was confirmed according to pathological evidence. Tissues were snap-frozen in liquid nitrogen and stored at − 80 °C before detection was performed. This study was approved by the Ethics Committee of Longhua Hospital (2019LCSY020), and informed consent was obtained from all participants.

M6A level was assayed using an RNA m6A quantification kit (ab185912, Abcam, USA) per the protocol used in a previous study [26]. Briefly, 200 ng of RNA was incubated for 60 min with capture antibody, after which detection antibody and enhancer solution were added. Finally, the samples were incubated with developer solution for 10 min. Absorbance was detected at a wavelength of 450 nm.

Tissue samples from the normal, adenoma, and CRC groups were fixed and then cut into 4-μm sections for immunohistochemistry (IHC). Tissue microarrays (TMAs) were obtained from Shanghai Outdo Biotech Co., Ltd. (Shanghai, China), and IHC was performed. In brief, tissue samples were treated with EDTA, as an antigen retrieval solution, and then incubated with m6A antibodies (1:100, 56,593, CST, USA) and METTL3 antibodies (1:500, ab195352, Abcam, USA) overnight at 4 °C. Subsequently, secondary antibodies were incubated for 1 h at 37 °C. Finally, the samples were stained and imaged. The scores for staining intensity were determined according to a staining intensity scale that ranged from 0 to 3+ points (0 for no staining, 1+ for weak immunoreactivity, 2+ for moderate immunoreactivity, and 3+ for strong immunoreactivity). Positive percentage was scored as follows: 0 for negative cells, 1+ for 1–25%, 2+ for 26–50%, 3+ for 51–75%, and 4+ for 76–100%. The intensity and positive proportion scores were then multiplied to obtain a composite score, which recorded as the scores of IHC. The composite score ranged from 0 to 12; a below average score indicates low expression, whereas an above average score indicates high expression. The clinical characteristics of the samples are summarized in Additional file 4: Tables S2 and S4.

The m6A antibody (1:300, 56,593, CST), METTL3 antibody (1:1000, ab195352, Abcam), and CRB3 antibody (1:500, PA5–53092, Thermo Fisher Scientific, USA) were used. Samples were incubated with m6A, METTL3, and CRB3 antibodies overnight at 4 °C. Subsequently, secondary antibodies conjugated with Alexa Fluor were incubated for 1 h at 37 °C. Finally, the samples were stained and imaged.

Nuclear extracts of adenoma and CRC tissues were performed using a nuclear extraction kit I (OP-0002, Epigentek, USA). Briefly, tissues were cut into small pieces, and homogenized. Then they were incubated on ice for 15 min and centrifuged. Finally, the supernatant was collected. Then m6A methylase activity was performed using m6A methylase activity/inhibition assay kit (P-9019, Epigentek, USA) per the protocol. In brief, 80 μl of binding solution and 2 μl of m6A methylase substrate were added into each strip well, which were incubated at 37 °C for 90 min. Then binding solution was removed, and 46 μl of working methylase buffer and 4 μl of nuclear extracts were added into each well. Fifty microlitre of the diluted capture antibody were added and signal detection was performed at 450 nm. Finally, m6A methylase activity was calculated per the protocol.

Normal colon cells (FHC) and CRC cells (HCT116, HT29, SW480, and SW620; Shanghai Cell Bank, Shanghai, China) were cultured in Dulbecco’s Modified Eagle Medium supplemented with 10% fetal bovine serum (FBS) and penicillin/streptomycin (100 U/mL) in an incubator with 5% CO₂ at 37 °C. In addition, 293 T cells obtained from the American Type Culture Collection were cultured in Roswell Park Memorial Institute 1640 Medium with 10% FBS and penicillin/streptomycin (100 U/mL; Gibco, USA). METTL3 short hairpin RNA plasmid, CRB3 short hairpin RNA plasmid, or negative control (Genomeditech, China) were transfected using FuGene HD transfection reagent (E2311, Promega, USA) per the protocol used in a previous study [27].

After the transfection of HCT116 and SW620 cells was completed, these cells were seeded on 96-well plates at a concentration of 2 × 10⁴ cells and cultured for 0, 24, 48, and 72 h. Subsequently, 10-μL cell counting kit-8 (CCK8) was added to each well. After incubation at 37 °C for 1 h, the absorbance value was detected at 450 nm.

After the transfection of HCT116 and SW620 cells was completed, these cells were seeded on a six-well dish with a culture insert (ibidi, Germany) at a concentration of 3 × 10⁴ cells. After 24 h, the culture insert was removed, and the cells were washed twice with phosphate buffer. Thereafter, 2-mL serum-free medium was added to each dish, which was then set aside for 48 h. Images were captured, and the wound area was measured using ImageJ software (National Institutes of Health, USA).

Six-well plates with 8-μm chambers (Corning, USA) were used to assess cellular migration (without Matrigel) or invasion (with Matrigel). In brief, transfected HCT116 and SW620 cells were seeded in six-well plates at a concentration of 1 × 10⁵ cells. In total, 200-μL serum-free medium was added to the upper chamber, and 600-μL of medium with 30% FBS was added to the lower chamber, and they were set aside for 48 h. The cells were then fixed with 4% paraformaldehyde for 30 min and stained with 0.1% crystal violet solution for 15 min. Five fields were randomly selected to calculate the area of migrating or invading cells.

Total RNA was extracted using TRIzol reagent (Ambion, USA). Complementary DNA was synthesized using an EVM-MLV reverse transcription kit (Aikeri Biotech, China). The amplification reaction was performed using the SYBR-Green quantitative real-time polymerase chain reaction (qPCR) kit (Thermo Fisher Scientific, USA). Gene expression was normalized using β-actin. The primers are listed in Additional file 4: Table S1.

Cells were collected and lysed, and protein concentration was determined. The protein was separated and transferred to a polyvinylidene difluoride membrane followed by incubation with 5% milk at room temperature for 1 h. The membrane was incubated at 4 °C overnight with the following antibodies: METTL3 (1:1000, 86,132, CST), CRB3 (1:1000, NBP1–98328, Novus Biologicals, USA), YTHDF2 (1:1000, 80,014, CST), macrophage stimulating 1 (MST1, 1:1000, 3682, CST), phospho-MST1 (1:1000, 49,332, CST), salvador family WW domain containing protein 1 (SAV1, 1:1000, 13,301, CST), large tumor suppressor kinase 1 (LATS1, 1:1000, 3477, CST), phospho-LATS1 (1:1000, 8654, CST), MOB kinase activator 1 (MOB1, 1:1000, 13,730, CST), phospho-MOB1 (1:1000, 8699, CST), Yes1-associated transcriptional regulator (YAP, 1:1000, 4912, CST), phospho-YAP (1:1000, 13,008, CST), histone (1:1000, 4499S, CST), and β-actin (1:1000, 4970 s, CST). The secondary antibody was then added, and the membrane was incubated at room temperature for 1 h; protein expression was observed using a chemiluminescence gel imaging system (Tanon 5200, China).

Total RNA was quantified using the NanoDrop ND-1000. The sample preparation and microarray hybridization were performed per Arraystar’s standard protocols. In brief, the total RNAs were immunoprecipitated with anti-m6A antibody. The modified RNAs were eluted from the immunoprecipitated magnetic beads as the “IP.” The unmodified RNAs were recovered from the supernatant as “Sup.” The IP and Sup RNAs were labeled with Cy5 and Cy3, respectively, using the Arraystar Super RNA Labeling Kit. The RNAs were combined and hybridized onto an Arraystar Human m6A Epitranscriptomic Microarray (8x60K, Arraystar). After the slides were washed, the arrays were scanned using an Agilent Scanner G2505C.

The m6A level of CRB3 was determined using a methylated RNA immunoprecipitation-qPCR assay per the protocol used in previous studies [4, 14]. In brief, 100 μg of total RNA was purified using an mRNA purification kit (61,006, Invitrogen, USA) and subsequently sonicated. And 2 μg of the RNA was saved as the input. These RNA fragments were further immunoprecipitated with 5 μg of anti-m6A antibody (202,003, Synaptic Systems). Samples were incubated using protein A/G magnetic beads (88,803, Thermo Fisher Scientific, USA). The enriched RNA was analyzed through qPCR, and the m6A enrichment was normalized using input. The primers are listed in Additional file 4: Table S4.

HCT116 cells were seeded on 6-well plates for 24 h, and then treated with 5 μg/mL actinomycin D (MCE, USA) at the 0, 2, 4, 8, 24 h. Total RNA was then isolated by using TRIzol (Ambion) and analyzed through qPCR. The mRNA expression for each group at a given time was calculated and normalized using β-actin.

RNA immunoprecipitation (RIP) assay was performed using the Magna RIP RNA-Binding Protein Immunoprecipitation Kit (17–700, Millipore, USA) per the manufacturer’s instructions. In brief, the cell samples were collected and incubated using magnetic beads with 5 μg of mouse immunoglobulin G (17–700, Millipore) or YTHDF2 (80,014, CST). Then protein-RNA complexes were then precipitated with a magnetic separator, and the supernatant was discarded. Thereafter, each immunoprecipitate was incubated with proteinase K, and the RNA was purified. Finally, the relative interaction between YTHDF2 and CRB3 transcripts was determined through qPCR and normalized to the input.

To perform this assay, 293 T cells were seeded on 24-well plates and transfected using FuGene HD transfection reagent (E2311, Promega) per the protocol used in a previous study [28]. In brief, 293 T cells were transfected with CRB3 wild type (CRB3-WT; Genomeditech) or CRB3 variant (CRB3-Mut; Genomeditech) plasmid with or without METTL3 knockdown for 6 h. Luciferase activity was detected using the dual-luciferase reporter system (E1910, Promega).

A study indicated that approximately 85% of CRCs may have transformed from adenomas [29]; therefore, we examined the m6A levels in both adenoma and CRC tissues. The results revealed substantially higher m6A levels in both adenoma and CRC groups relative to the normal group (Fig. 1a–c). A TMA was performed to explore the correlation between m6A and survival in patients with CRC. The results indicated a substantially higher m6A level in the CRC group relative to the normal group (Fig. 1d and e, Additional file 4: Table S2). Furthermore, patients with CRC with high m6A level exhibited shorter overall survival (Fig. 1f, Additional file 4: Table S3).

M6A modification is regulated by RNA methyltransferase and demethylase, hence we assessed the expression of methyltransferase (METTL3, METTL14 and WTAP) and demethylase (FTO and ALKBH5) in both adenoma and CRC. The results indicated that the protein level of METTL3 was significantly higher in the CRC group relative to the normal and adenoma groups (Fig. 2a–c); this finding is consistent with the results in the database (Additional file 1: Fig. S1a). And mRNA level of METTL14 was significantly higher in the adenoma group but not in the CRC group (Additional file 1: Fig. S1c). Relative to the normal group, FTO was significantly lower in the adenoma group but not in the CRC group (Additional file 1: Fig. S1d). WTAP and ALKBH5 levels did not differ significantly among the three groups (Additional file 1: Fig. S1e and 1f). M6A methylase activity was significantly higher in the CRC group relative to the adenoma group (Fig. 2d). In CRC cells, the mRNA and protein levels of METTL3 were also substantially higher than in normal colon cells (Fig. 2e and f). To explore the correlation between METTL3 and survival of patients with CRC, a TMA was performed. The results indicated that the level of METTL3 was substantially higher in the CRC group relative to the normal group (Fig. 2g and h, Additional file 4: Table S4). The patients with CRC with high METTL3 level also exhibited shorter overall survival relative to those without a high METTL3 level (Fig. 2i, Additional file 4: Table S5), which suggests that METTL3 level can serve as a prognostic marker of CRC.

To investigate the function of METTL3 in CRC, METTL3 knockdown was performed in HCT116 and SW620 cells (Fig. 3a), and it significantly inhibited the proliferation of HCT116 and SW620 cells (Fig. 3b). Transwell assays indicated that METTL3 knockdown substantially inhibited the migration and invasion of HCT116 and SW620 cells (Fig. 3c and d), and METTL3 knockdown also significantly reduced the migration speed of HCT116 and SW620 cells (Fig. 3e and f).

To investigate the potential mechanism of METTL3 in CRC progression, an epitranscriptomic microarray was performed, and the data obtained were analyzed. For m6A methylation, 363 differentially m6A-methylated sites (DMSs) were identified; they comprised 353 upregulated DMSs and 10 downregulated DMSs (Fig. 4a, Additional file 5). A further analysis revealed that these 363 DMSs belonged to 349 differentially m6A-methylated genes (DMGs), which comprised 341 upregulated DMGs and 8 downregulated DMGs (Fig. 4b). For m6A quantity, 248 DMSs were identified; they comprised 191 upregulated DMSs and 57 downregulated DMSs (Fig. 4c, Additional file 6). A further analysis revealed that these 248 DMSs belonged to 224 DMGs, which comprised 175 upregulated DMGs and 49 downregulated DMGs (Fig. 4d). Subsequently, main functions were identified using the Gene Ontology and Kyoto Encyclopedia of Genes and Genomes, which contain sequence-specific DNA binding, RNA polymerase II transcription factor activity, Toll-like receptor binding, beta-alanine metabolism, alpha-linolenic acid metabolism, and nicotinate and nicotinamide metabolism (Fig. 4e-h).

Seven overlapping DMGs between m6A methylation and quantity were filtered using a Venn diagram (Fig. 5a), and a qPCR analysis indicated a substantial increase in CRB3 level after METTL3 knockdown (Fig. 5b and c). The results obtained from The Cancer Genome Atlas (TCGA) database revealed a substantial downregulation of CRB3 level in the CRC group (Additional file 2: Fig. S2b). A survival analysis indicated that the patients with CRC with high CRB3 level exhibited higher overall survival and disease-free survival relative to those without high CRB3 level (Additional file 2: Fig. S2c). Although METTL3 knockdown also promoted bridging integrator 1 (BIN1) expression, a substantial increase in BIN1 expression was observed in the CRC group (Additional file 2: Fig. S2a and 2d). The survival analysis did not indicate a correlation between BIN1 level and survival in patients with CRC (Additional file 2: Fig. S2e). In addition, the m6A level of CRB3 was substantially reduced after METTL3 knockdown (Fig. 5d). Luciferase reporters were revealed to determine the effect of m6A modification on CRB3 expression. For the variant form of CRB3, the adenosine bases in m6A consensus sequences (GGAC) were replaced by cytosine; thus, m6A modification was abolished. The results indicated that the transcriptional level of wild-type CRB3 significantly increased after METTL3 knockdown but not its level of variation (Fig. 5e). An RNA stability assay revealed that METTL3 knockdown substantially inhibited the degradation of CRB3 mRNA (Fig. 5f). Moreover, data from the TCGA database indicated that METTL3 mRNA expression in CRC tissues was negatively associated with CRB3 levels (Fig. 5g). Studies have reported that YTHDF2 can target mRNAs by recognizing m6A motif in CRC [5, 12]; thus, we explored the effect of YTHDF2 on CRB3. The results indicated that the expression of YTHDF2 was also significantly higher in adenoma and CRC tissues relative to normal tissues (Additional file 3: Fig. S3), and YTHDF2 knockdown substantially increased the level of CRB3 (Fig. 5h and i). RIP assays also verified the direct interaction between the YTHDF2 and CRB3 mRNA, and this direct interaction was impaired after METTL3 inhibition in SW620 and HCT116 cells (Fig. 5j and k). The results indicated that METTL3 regulated the expression of CRB3 in an m6A-YTHDF2-dependent manner.

Our results also indicated that CRB3 level was substantially lower in the adenoma and CRC groups than in the normal group, which was consistent with the data from TCGA database (Fig. 6a and b, Additional file 2: Fig. S2b). To further investigate the function of CRB3 in CRC, CRB3 knockdown was performed in HCT116 and SW620 cells. CRB3 knockdown significantly promoted the proliferation, migration and invasion of HCT116 and SW620 cells (Fig. 6c-e). CRB3 knockdown also significantly increased the migration speed of HCT116 and SW620 cells (Fig. 6f and g). These results verified that CRB3 regulated CRC progression.

Studies have demonstrated that CRB3 can regulate the Hippo pathway [30, 31]. In the present study, we discovered that CRB3 knockdown reduced MST1, LATS1, MOB1, and YAP phosphorylation levels, and it reduced the SAV1 levels in HCT116 and SW620 cells (Fig. 7a and b). Conversely, METTL3 knockdown substantially increased MST1, LATS1, MOB1, and YAP phosphorylation levels in HCT116 and SW620 cells (Fig. 7c and d). Studies have demonstrated that YAP enters the nucleus and acts as an oncogene; thus, we detected the level of YAP in nuclei. The results indicated that CRB3 knockdown markedly increased YAP protein levels in the nuclei of HCT116 and SW620 cells (Fig. 7e and f); however, METTL3 knockdown substantially reduced the YAP protein levels in the nuclei of HCT116 and SW620 cells (Fig. 7g and h).

To test whether CRB3 functions downstream of METTL3, we knocked down CRB3 in the METTL3 knockdown background. The results revealed that the effects of METTL3 knockdown on cell proliferation, migration, and invasion were rescued by the CRB3 knockdown (Fig. 8a-d). In addition, CRB3 knockdown reduced the MST1, LATS1, MOB1, and YAP phosphorylation levels caused by METTL3 knockdown, indicating that CRB3 knockdown repressed activation of hippo pathway that was caused by METTL3 knockdown (Fig. 8e and f).